Computer-implemented method for reducing the noise signature of a platform-mounted propulsion system, data processing system, computer program, computer-readable medium, propulsion system and watercraft

Abstract

The underlying invention relates in particular to a method for reducing the noise signature of a drive system mounted on a platform (1), which comprises at least one driving and one driven (2) component. A control signal with at least one setpoint signal (S) is provided for operational control. The method is carried out iteratively, with each iteration step (i) comprising the following steps: a. Generating a modulated setpoint signal (mS), b. Applying the modulated setpoint signal (mS) instead of the setpoint signal (S), c. continuous recording of vibration signals (V), d. sliding determination of vibration spectra with sliding window from the vibration signals (V), and determination of frequency components (f R , 5f R , ...) for each window from respective vibration spectra, e. Calculating a cost function (K) for the respective frequency components (fR, 5fR, ...), and determining the window for which the cost function (K) has a minimum, f. Calculating a time average (ZM) of the modulation signal (M) for the window associated with the minimum, and g. Correction of the setpoint signal (S) based on the time mean (ZM) and application of the corrected setpoint signal (S') instead of the setpoint signal (S).

Description

[0001] The underlying invention relates to a computer-implemented method for reducing the noise signature in a propulsion system mounted on a platform, a data processing system, a computer program, a computer-readable medium, a propulsion system, in particular for watercraft, and a watercraft, in particular a ship.

[0002] For drive systems with one or more driving and one or more driven components mounted on or coupled to a platform, vibration behavior and noise signature are sometimes crucial operating characteristics. Therefore, for many drive systems, it is desirable to optimize the noise signature and vibrations during operation, particularly to reduce them or minimize them.

[0003] Vibrations and the noise signature can result, for example, from the operation of a drive motor or a driven component as such, as well as from interference or dynamic coupling or superposition of individual operating frequencies of the respective components. In particular, dynamic coupling and interference, also known as beats, can contribute significantly to the noise signature in systems with multiple driving or driven components under certain operating conditions, for example, in the form of a low-frequency hum.

[0004] Since the noise signature emanating from a drive system always affects at least the immediate surroundings or environment, it is desirable, if not even required by law, to reduce the noise signature or to reduce it as far as possible.

[0005] Particularly in systems with two or more driving or driven components such as wings or blades, such as in wind turbines, ship propulsion systems, turbo compressors, engines, generators, etc., dynamic couplings can lead to undesirable noise emissions that affect the environment and are accordingly perceptible or detectable.

[0006] For example, the noise signature on ships significantly impairs on-board comfort and leads to noise pollution for underwater wildlife. Furthermore, a pronounced noise signature increases the acoustic detection probability of ships in a military context. The same applies to other systems with a pronounced noise signature.

[0007] The noise signature on ships, for example, can be caused by a variety of sources or objects, such as driving noise due to water displacement, diesel engine hum, electric motors, power electronics, compressors, pumps, fans, etc. There are a variety of measures to reduce the noise signature.

[0008] One way to reduce the noise signature is to consider and optimize the noise of a single noise-generating object, such as a motor, a diesel engine, a pump, etc. Examples of single noise optimization include sound-dynamically decoupled assembly, layout and design of the respective objects, and / or the selection of a noise-reduced operating mode, and sound insulation. Single optimizations can, for example, concern the design of the ship's shape / hull, the propeller or blades, the engine(s), and / or the coordination between the engine and the power control system.

[0009] With regard to possible improvements for reducing noise generation and underwater sound based on singular system component considerations, reference is made to Tom A. Smith et al., “Underwater radiated noise from marine vessels: A review of noise reduction methods and technology”, Ocean Engineering, Vol. 266, Part 1, Dec. 15, 2022.

[0010] Individual components on the ship and other systems form a system network, which are more or less strongly coupled to each other depending on the object type, function, and placement, thus forming a dynamic system. Dynamic couplings can lead to high levels of noise being perceived / measured under certain operating conditions, even though the respective components operate at minimal noise levels on their own.

[0011] On a ship, one subsystem of the overall ship system that can cause such behavior can include the propeller-drive shaft-electric motor-drive system. Such a drive system can be present multiple times on ships, so there can be a multitude of dynamic coupling paths between these drive systems. Examples include hydrodynamic couplings (e.g., jet direction in nacelle propellers or Voith-Schneider propellers), imbalances in the drive trains, inner-propeller blade couplings, which can lead to vibrations and noise and be transmitted to other propellers, etc.

[0012] On ships with multiple propellers (including aircraft propellers and other systems), specific relative propeller rotor positions (relative phase positions of the propellers to each other) can exist at which noise levels can be significantly reduced. Such rotor positions result in, among other things, destructive superposition and cancellation effects of sound and structure-borne sound waves, which can improve noise levels. The achievable amount of noise improvement generally depends on the position of the measuring / perception points as well as the position of the noise-generating sources.

[0013] In addition, an optimal relative rotor position for ship propellers, where "optimal" is understood here as the minimum achievable noise level at a specific location or area, is usually dependent on several factors, such as the position of the propellers on the ship and their relative alignment to one another, the ship's speed, currents, propeller type (e.g. controllable pitch propeller, propeller with constant pitch), thrust and / or thrust direction of the individual propellers, speed of the propellers and pitch, sea state, ship loading, draught, etc.

[0014] If, for example, in a propeller-drive shaft-electric motor combination on a ship, the relative rotor positions to each other are not controlled, small speed differences typically occur between the propellers. These differences result in a continuous change in the relative rotor position of the propellers. This results in a beating of the noise amplitudes on the ship, which is also referred to in the literature as "vibration beating."

[0015] For example, with regard to relative rotor position control – also known as synchrophasing – on ships, reference is made to “Synchrophasing – Enhanced vibration performance for twin propulsion vessels”, MAN Energy Solutions, 5510-0247-00 Mar 2020, available at: https: / / www.man-es.com / docs / defaultsource / marine / tools / synchrophasing.pdf?sfvrsn=3c464a88_14. In this approach, a ship is operated at a constant speed and under constant environmental conditions with slightly different propeller speeds during a test run, so that the relative rotor position of the propellers changes comparatively slowly. Parallel to this operation, vibrations and noise are measured in certain areas of the ship, and a vibration and noise background is determined as a function of the relative rotor position of the propellers. The vibration and noise background of the test run is then used to determine the optimal relative rotor position (with, for example, min.A noise signature is determined, which is then applied during real operation of the ship via a relative rotor position control. However, the determined optimal relative rotor position is only optimal for the respective operating situation investigated during the test run, i.e., for the test operation, and can lead to a suboptimal noise level in real operation with changed ambient conditions and operating parameters (ship load, draft, etc.). Furthermore, the number of test runs is comparatively small, as they are complex and time-consuming, so that not all possible operating situations can be simulated, and operating situations with suboptimal noise levels can occur in real operation.

[0016] The optimization of the noise level and its disadvantages described above for ships apply analogously to other systems, such as wind turbines, turbo compressors, etc. with, for example, several driven and driving components.

[0017] Based on this, it is an object of the invention to provide a novel method for reducing the noise signature of a propulsion system operated on a platform. In particular, an improved method for optimizing the noise level is to be provided. With a similar objective, a data processing system, a computer program, a computer-readable medium, a propulsion system, in particular for watercraft, and a watercraft are also to be provided.

[0018] This object is achieved by the features of the independent claims. Further developments emerge from the subclaims and the following description of developments and exemplary embodiments.

[0019] According to one embodiment, a computer-implemented method for reducing the noise signature of a drive system mounted on a platform is provided. The drive system comprises at least one driving and one driven component. The driving component can be, for example, an engine, a diesel engine or electric motor, or a drive blade or rotor, e.g. of a wind turbine or a wind turbine. The driven component can be, for example, a rotor, such as a ship's rotor, in particular a so-called Voith-Schneider propeller, a compressor wheel of a turbo compressor, or a generator of a wind turbine or a wind turbine. In embodiments, in particular, two independently driven or driving components can be present, which are optimized with regard to the noise signature.

[0020] In embodiments, the drive system can comprise two, in particular exactly two, or more than two driving components or drives, such as diesel or electric motors, and two, in particular exactly two, or more than two propellers or rotors, each drive-coupled to one of the drives. A propeller or rotor can comprise two or more propeller or rotor blades, as is the case, for example, with a ship's propeller, in particular a Voith-Schneider propeller. In embodiments such as a wind turbine, the propeller or rotor - also called a drive rotor - can form a driving component, and a generator connected to the propeller or rotor can form a driven component.

[0021] However, the number of driving or driven components and their number in relation to one another is not limited within the scope of the invention, ie the method proposed herein is suitable for essentially any combination of driving and driven components.

[0022] According to the method, a control signal or a setpoint signal specification with at least one setpoint signal is provided for controlling the operation of the driving or driven component according to an operating requirement. The setpoint signal can be, for example, a signal for controlling the speed of the driving component and / or the relative rotor position of the driven component. However, the control signal or setpoint signal is not limited to the aforementioned possibilities and can include other setpoint operating parameters for the at least one driving or driven component.

[0023] The driving component may comprise a drive motor, for example an electric motor, such as a permanent magnet synchronous motor (PMSM) or be designed as such.

[0024] In particular, when the method is applied to a ship's propulsion system with, for example, two drive trains, each comprising a drive motor and a propeller or rotor coupled to it for drive purposes, the setpoint signal can be a setpoint specification for the relative rotor position of the rotors. In the case of a wind turbine or a wind turbine, the setpoint signal can, for example, be a setpoint specification for the relative rotor position of multi-bladed drive rotors, with the drive rotor forming a driving component. If, in a wind turbine with several wind turbines, for example, the relative rotor position of the generators coupled to corresponding drive rotors is controlled, the setpoint signal accordingly relates to the driven component, i.e. the generator. In the case of a turbocompressor, the driving component can be the compressor drive and the driven component the compressor rotor.

[0025] The proposed method is carried out iteratively, with each iteration step comprising the following steps: a. generating a modulated setpoint signal by modulating the setpoint signal, in particular the setpoint signal present or valid at the time, in particular the setpoint signal present at the start of the iteration, with a modulation signal that can be described by a periodic function, such that a time average of the modulated setpoint signal corresponds, in particular substantially, to a time average of the setpoint signal; b. Applying the modulated setpoint signal instead of the setpoint signal to control the operation of the driving or driven component, c. during the application of the modulated setpoint signal, continuously detecting vibration signals representative of local vibrations of the platform and / or the component by at least one vibration sensor, d. smoothly determining vibration spectra with a sliding window from the vibration signals, wherein the sliding window has a predetermined window length which is smaller than a period of the modulation signal, and determining frequency components for each window from respective vibration spectra; e. for each of the windows, calculating a cost function for the respective frequency components, and determining the window for which the cost function has a minimum; f. Calculating a time average of the modulation signal for the window associated with the minimum; and g. Correction of the setpoint signal based on the time average, for example by adding the time average and applying the corrected setpoint signal instead of the setpoint signal.

[0026] The time averages for the setpoint signal and the modulated setpoint signal preferably relate to the length or duration of the respective iteration interval or iteration step, or to the duration of the acquisition and continuous determination of the vibration spectra. The modulation preferably takes place in such a way that the time average of the setpoint signal and the time average of the modulated setpoint signal are equal or essentially equal in the respective relevant time interval, e.g., the duration of the acquisition and continuous determination of the vibration spectra.

[0027] The setpoint signal can be specified, for example, by a setpoint signal specification, or result from a corrected setpoint signal from one or more previous iterations.

[0028] Preferably, the duration of the acquisition of the vibration spectra and the smooth determination of the vibration spectra (step c. and partly step d.) corresponds to a period length of the periodic function.

[0029] The modulated setpoint signal can be generated, for example, by superimposing or modulating the setpoint signal with a modulation function, such as a periodic modulation function whose time integral over the iteration interval is zero. Suitable modulation functions include sine or cosine functions. The terms modulation signal and modulation function are used synonymously herein unless explicitly stated otherwise.

[0030] According to step b, the modulated setpoint signal is applied to the driving or driven component to control its operation. If the setpoint signal is intended for the driving or driven component, the modulated setpoint signal is applied to the driving or driven component accordingly.

[0031] According to step c., vibration signals are recorded in the respective iteration step when the modulated setpoint signal is applied. If the setpoint signal relates to the relative rotor position, e.g., of two rotors driven by a drive motor, the modulated setpoint signal causes a modulation of the relative rotor position. Based on the modulation of the relative rotor position and the continuous recording of the vibration signals, the noise signature can be determined, in simple terms, as a function of the modulation, and the relative rotor position present at the beginning of the iteration can then be changed or corrected, e.g., in the direction of the modulated relative rotor position, for which an improved or optimized noise signature results. As already mentioned at the beginning, "optimal" or "optimized" should be understood in connection with the noise level in the sense of a minimum achievable noise level, e.g.at one or more specific locations or areas, for example, the platform. If the value of the non-modulated setpoint signal, which is always passed through during modulation, provides the optimal noise level for the iteration interval, the setpoint signal or the value of the non-modulated setpoint signal is used as the corrected setpoint signal in accordance with step g).

[0032] The vibration signals can be detected by one, two, or more vibration sensors, such as piezoelectric vibration sensors, mounted locally at predefined or selected locations on the platform and coupled to them. It is also possible for one or more vibration sensors to be coupled to platform components, which in turn are coupled to the platform. In addition to vibration sensors that can detect vibrations through vibration coupling or mechanical coupling to the platform or a component of the platform, such as piezo sensors, it is also possible to use other types of sensors, such as optical sensors, acceleration sensors, and force sensors (e.g., strain gauges, etc.), to detect signals representative of the noise signature.

[0033] In step d., vibration spectra are determined continuously using a sliding window or a sliding time window or a rolling time window, whereby "sliding determination" should be understood in the usual mathematical sense. In particular, a sliding determination should be understood as the determination of vibration spectra from vibration signals that were continuously recorded within a respective, predetermined period of time, i.e., the window or time window with a predetermined, fixed window length. For example, the vibration spectra can be determined by applying a sliding Fourier transform (FT) or a sliding fast Fourier transform (FFT) to the recorded vibration signals.

[0034] Step e. comprises calculating a cost function for each of the windows, i.e. an evaluation of the respective vibration spectrum specific to each window. A cost function should be understood in particular as a function which maps the frequency components determined in step d. or values, quantities or vectors determined therefrom to a value or vector which is characteristic of the vibration behavior or vibrations, in particular the noise signature, of the drive system or the platform, or describes the vibration behavior, etc. The term “cost” in the sense of the invention therefore relates to vibrations in the sense of the noise signature, which are to be optimized or minimized in the context of the invention. For example, the cost function can comprise a weighted or unweighted sum or quadratic sum of components of a respective vibration spectrum.

[0035] In step f., the time average of the modulation signal is calculated for the window for which the cost function is minimal within an iteration, i.e. for which optimal (minimal) vibration behavior is present within the iteration. The time average of the modulation signal in the respective window corresponds to the average deviation of the modulated setpoint signal from the setpoint signal present at the start of the iteration. The invention is based on the finding that the positive or negative deviation from the setpoint signal present at the start of the iteration corresponding to the time average specifies - in short - the "direction" in which a change in the setpoint signal leads to an optimization or minimization of vibrations and thus of the noise signature. Accordingly, in step g., the setpoint signal present at the start of the iteration is corrected on the basis of the time average.For example, it is possible to correct the setpoint signal by adding the time average to the setpoint signal. For the purposes of the application, addition also includes subtraction, which represents the addition of a negative value. A correction of the setpoint signal based on the time average can further comprise multiplying the setpoint signal by a multiplier or factor, in particular by a stretching or compression factor. The multiplier can be determined from the time average, in particular such that the correction effects a change in the setpoint signal toward a reduced noise signature. For example, the multiplier can be determined such that the product of the multiplier and the setpoint signal corresponds to the sum of the setpoint signal and the time average.

[0036] Overall, the above explanations demonstrate that the steps performed during iteration enable optimization or minimization of the noise signature within an optimization process. In particular, the iterative and continuous correction of the setpoint signal enables optimization of the setpoint signal toward an optimized or minimized noise signature. The iterative process makes it possible to optimize operating parameters present at the beginning of each iteration, e.g., the setpoint signal, with regard to the noise signature. The continuous execution of iteration steps can be viewed as a control loop. Thus, the iteration represents the structure and mode of operation of a control loop.

[0037] The method can be carried out continuously during operation of the drive system and results in an optimization, minimization, or reduction of the noise signature. The special feature of the proposed method is that the noise signature can be optimized, minimized, or reduced during operation even without prior test runs or test drives. A reduction in the noise signature can be carried out dynamically, for example in real time. Optimization of the noise signature in real time is possible in particular due to a deterministic response time of the iteration, whereby the response time is given or defined by the length of the iteration interval. The reduction or optimization of the noise signature can be carried out parallel to normal operation, whereby existing, real-world framework conditions, operating states, and external circumstances can be taken into account.The latter is not possible with state-of-the-art methods, in which the optimization of the noise signature is based on test runs or test drives, because it is difficult to run through or simulate all possible operating conditions during test runs and test drives. In this respect, the novel method proposed here offers, in particular, efficient optimization of the noise signature in drive systems under real operating conditions.

[0038] According to embodiments, an iteration of steps a. to g. can be terminated and continued with a new iteration using an adapted modulation function if, for example, the setpoint signal changes according to a setpoint signal specification by a higher-level control unit. This means that the iteration can be adapted to the respective new operating situation, i.e., the new setpoint signal specification. In other words, the method is also suitable for applications in which the setpoint signal can be changed by a setpoint signal specification.

[0039] According to embodiments, the period of the modulation signal and / or an amplitude of the modulation signal can be varied for successive iteration steps. Preferably, the period of the modulation signal and / or an amplitude of the modulation signal is / are calculated as a function of the setpoint signal and / or a rotational speed applied, determined, measured or estimated in a previous or respective iteration step. At the start time of an iteration, for example upon initiation of the method or a further iteration after completion of a previous iteration, currently available setpoint signals or control signal values ​​and, if required according to embodiments, further operating parameters, such as the rotational speed of the driving or driven components, can be read in, measured or estimated, which then form the basis or starting point for parameterizing the periodic function orof the modulation signal or the modulation function, and the window length. In particular, an estimate of quantities or parameters required for a respective iteration and parameterization can be made using an observer, in particular a Luenberger observer. For example, the speed of the driving or driven component can be estimated on the basis of or using a Luenberger observer. Overall, the parameters required to execute an iteration step, in particular parameters for the modulation function, such as the amplitude and period of the modulation function, can be redetermined at the start time of a new iteration. This enables, for example, a reliable optimization of the noise signature, because each newly determined parameter characterizes the respective real operating state.

[0040] According to embodiments, it can be provided that the window length is proportional to the period of the modulation signal and / or that the window length is inversely proportional to the number of revolutions of the driving or driven component per period and / or that the window length is inversely proportional to a rotational speed of the driving or driven component in the respective iteration step. In particular, the period of the modulation signal or the modulation function can be an integer multiple of the window length. In embodiments, at the beginning of a new iteration step, the period can be derived from mean values ​​of the estimated / measured rotational speeds of the driving components or driven components, in particular the rotors. The rotational speed can, e.g.Depending on the application, it can be derived from a control signal for the driving or driven component or determined by measuring or estimating rotor position sensors, e.g. on rotors.

[0041] The functional relationships for the window length mentioned above have proven to be particularly advantageous and efficient for optimizing, in particular reducing or minimizing, the noise signature.

[0042] According to embodiments, the modulation signal or the modulation function can be described by a sine or cosine function or can be given by a sine or cosine function. These functions are particularly suitable for modulation because, when modulated onto the setpoint signal determined or present at the beginning of an iteration step, they do not change, or essentially do not change, the time average over each complete period due to the periodicity.

[0043] According to embodiments, a fundamental frequency and harmonic overtones or harmonics of the fundamental frequency can be determined as frequency components in step d. In step e., the cost function can be calculated as a weighted sum, in particular a weighted quadratic sum, of the fundamental frequency and the harmonics. Taking harmonics into account enables efficient optimization, reduction, or minimization of the noise signature, because harmonics in particular can significantly influence the noise signature depending on the situation. Taking harmonics into account, in particular, enables optimization of the noise signature based on real-world conditions.

[0044] According to embodiments, the at least one driving or driven component can comprise a rotor, for example a ship's rotor as the driven component, a multi-bladed drive rotor of a wind turbine as the driving component, a rotor of a motor as the driving component or a rotor of a generator as the driven component. In preferred embodiments, the setpoint signal preferably specifies a speed, a torque, a power and / or, more preferably, a phase position of the rotor or at least of a rotor blade connected to the rotor(s). In particular, in one embodiment, the speed can be realized with the manipulated variable. It is also possible to derive a power or torque manipulated variable from this manipulated variable and, for example, from the moment of inertia and the load behavior known for the drive system, which then serves as the basis for the iteration.In embodiments with several, in particular two, rotors as driven elements, the manipulated variable can define a relative rotor position of several, preferably two, rotors.

[0045] According to embodiments with a rotor, in particular as described above, the method can comprise at least one of the following features or conditions: h. The period of the modulation signal is inversely proportional to the speed of the rotor. i. The period of the modulation signal is proportional to the number of revolutions of the rotor during a period of, e.g., the previous iteration step and / or the modulation function. j. The window length in an iteration step is inversely proportional to the number of revolutions of the rotor during a period, in particular of the modulation function, of a respective / previous iteration step.

[0046] The speed of the rotor or the number of revolutions per period can be determined, e.g. estimated, etc., at the beginning of an iteration step, as described above.

[0047] Conditions h. to j. can be implemented under the assumption or boundary condition that the average speed of the rotor or propellers or the number of revolutions of the rotor is constant over one or more consecutive periods of the modulation signal or over the iteration step. Accordingly, the method can provide that a constant speed or number of revolutions is used as a basis for each or more consecutive periods. For each period or for several consecutive periods, the speed or the number of revolutions can therefore be determined and used to parameterize the modulation signal. As already mentioned, in embodiments, the speed or number of revolutions can be determined for a respective iteration step and assumed to be constant for the entire iteration step.However, it is also possible for the modulation signal and period duration to be re-parameterized within an iteration step in the event of a change in the speed or number of revolutions. For example, the speed or number of revolutions can be determined, in particular estimated, and / or determined based on the parameterization of a previous iteration step or a previous period duration.

[0048] The number of revolutions per period of the modulation signal can be specified as a fixed parameter, particularly in the context of determining the window length. For example, it is possible for the parameter to be specified by a user and / or changed during operation, e.g., by a control unit or control input. If the parameter is changed, the correspondingly changed parameter value can then be applied at the beginning of a new iteration. This allows, for example, a stretching or compression of the period. It is also possible for the number of revolutions to be determined based on a Fourier transformation.

[0049] The mentioned parameterizations of the period of the modulation function or the window length have proven to be advantageous with regard to an efficient optimization or reduction or minimization of the noise signature.

[0050] According to embodiments, it can be provided that the drive system comprises at least two driving or driven components, each having a rotor, wherein each rotor preferably comprises at least two rotor blades, and wherein the setpoint signal specifies a relative phase position for at least two of the rotors.

[0051] According to embodiments, it can be provided that the drive system comprises at least one, preferably at least two, driving components which are designed as a motor, in particular an electric motor. Furthermore, the drive system comprises two, in particular exactly two, rotors driven by the driving component(s) as driven components. Each rotor comprises a plurality of rotor blades connected to it. Preferably, the drive system is designed as a ship's propulsion system. A corresponding ship's propulsion system can in particular comprise two or more electric motor-driven Voith-Schneider propellers as driven components. In this case, the setpoint signal can specify a relative rotor position of at least two, preferably exactly two, of the rotors. For each time window, it is provided that an amplitude and the period of the modulation signal are determined. In preferred embodiments, each of the rotors can be driven by a respective associated motor orElectric motor, e.g. a PMSM, driven or drivable.

[0052] When applied to a ship propulsion system, as previously mentioned, it may be provided that k. the amplitude of the modulation signal is a function of the mean value of the rotor speeds, l. the period is inversely proportional to the average speed of the respective rotor, in particular in the respective / previous iteration step, and proportional to the number of revolutions of the rotor blades or the rotor per period, e.g. of the modulation signal, in particular of the respective / previous iteration step, and / or is inversely proportional to the average speed of the respective rotor, in particular in the respective / previous iteration step, m. the window length is proportional to the period, in particular of the modulation signal, and inversely proportional to the number of revolutions of the rotor blades per period, in particular of the respective / previous iteration step, and n. a fundamental frequency and integer multiples of the fundamental frequency multiplied by the number of rotor blades are taken into account as frequency components.

[0053] For example, corresponding values ​​determined in the previous iteration step or in the respective iteration step, e.g., estimated or read values, can be used as an average value of the rotational speeds or as the average rotational speed of the respective rotor or as the number of revolutions of the rotor blades per period. In particular, it is possible to estimate corresponding values ​​at the start of an iteration, analogous to the explanations above, and to use the estimated values.

[0054] Integer multiples of the fundamental frequency are, in particular, harmonics. Integer multiples of the fundamental frequency multiplied by the number of rotor blades can also be referred to as rotor blade frequencies and correspond to harmonics of the fundamental frequency. The fundamental frequency can, in particular, be the (rotational) frequency of the rotor.

[0055] Corresponding drive systems, particularly according to the embodiments described above, can comprise, for example, electric motor-driven ship propellers, in particular electric motor-driven Voith-Schneider propellers. The number of rotor blades or rotor vanes of a Voith-Schneider propeller can, for example, be odd. In particular, there can be three, five, seven, or nine rotor vanes. An even number of rotor vanes is also possible.

[0056] According to one embodiment, a data processing system is provided. The system comprises means, for example one or more control units, in particular processors, for executing the method steps according to a method embodiment described herein. The means are configured to receive or determine the setpoint signal and the vibration signals of the at least one vibration sensor and to provide the corrected setpoint signal for controlling the operation of the drive system or to control the drive system based on the corrected setpoint signal. For advantages and advantageous effects, reference is made to the explanations regarding the method embodiments.

[0057] According to one embodiment of the system, it may further comprise at least one vibration sensor, preferably a plurality of vibration sensors.

[0058] According to one embodiment, a computer program is provided which is set up or adapted such that, when executed on a data processing system as described above, it carries out or effects a method according to one of the method-related embodiments described herein.

[0059] According to one embodiment, a computer-readable medium is provided which comprises instructions which, when executed by a data processing system as described above, cause the system to carry out the method steps of a method according to one of the method-related embodiments described herein.

[0060] According to one embodiment, a propulsion system for watercraft is provided. The propulsion system for watercraft comprises at least one, preferably two motors, preferably designed as electric motors, as driving components, at least one, preferably two rotors as driven components and a plurality of rotor blades coupled to a respective rotor, one or more vibration sensors and a data processing system according to one of the embodiments described herein. The driven component preferably comprises, as the rotor with the rotor blades coupled thereto, an electrically driven Voith-Schneider rotor. In particular, the propulsion system can comprise at least two rotors, in particular Voith-Schneider rotors, each independently driven by a motor, in particular an electric motor.

[0061] According to one embodiment, a watercraft is provided comprising a platform with at least two rotors attached thereto, preferably independently driven by an (electric) motor, each having a plurality of rotor blades, in particular according to a Voith-Schneider rotor, one or, preferably, several vibration sensors coupled to the platform, and a control system for controlling the operation of the rotors. The control system comprises a data processing system according to one of the embodiments described herein. For advantages and advantageous effects, reference is made to the above explanations.

[0062] All embodiments described within the scope of the underlying invention are particularly applicable to propulsion systems with two or more than two Voith-Schneider propellers, each driven by an electric motor, e.g. a PMSM, and preferably designed for propulsion and / or directional control of a ship.

[0063] Exemplary embodiments and configurations of the invention are described below with reference to the accompanying figures. The embodiments and configurations are directed, among other things, to a ship propulsion system with two independently electric motor-driven Voith-Schneider Propellers (eVSP). However, as described above, the invention is also applicable to other systems, for example, wind turbines and turbocompressors, etc. In this respect, it is understood that the statements regarding the eVSP in connection with the figures are not limited to eVSP with regard to the applicability of noise signature optimization.

[0064] The figures show: Fig. 1 a flowchart of a method for optimizing the noise signature; Fig. 2 an example of a time sequence of a method for optimising the noise signature; Fig. 3 schematically shows the process of the sliding evaluation of vibration signals; Fig. 4 schematically shows an evaluation result of vibration signals based on an FFT; Fig. 5 schematically shows a structure of a ship platform with two eVSP; and Fig. 6 schematically shows a block structure of a propulsion system for a ship with two eVSPs and associated control and monitoring components.

[0065] Fig. Figure 1 shows a flow diagram of a method for optimizing the noise signature of a drive system with, for example, two eVSPs. The method is configured for iterative execution, where Fig. 1 the index “i” (i= 1, 2, 3, ...) denotes a respective iteration i.

[0066] In a first step a. of an iteration i, a modulated setpoint signal mS i generated. The corresponding setpoint signal S iFor example, it specifies the relative rotor position in an eVSP. The signal used to modulate the setpoint signal S i used modulation signal M i , hereinafter also modulation function M i is a periodic function, such as a sine function. The periodic function is chosen such that its time average cancels out for each period T, ie, is zero. Thus, when modulating the setpoint signal S i or when the setpoint signal S is superimposed i with the modulation signal M i or the modulation function M i the time average of the modulated setpoint signal mS i the time average of the setpoint signal S i In other words, the periodic modulation leads to a periodic change of the setpoint signal S i, for example, the relative rotor position, which, however, cancels out in the time average. Since the modulation cancels out in the time average, for example, when using a sine function as the modulation function M i , whose period is short compared to the observed operating period, it can be assumed that the operation of the drive system using the modulated setpoint signal mS i essentially corresponds to operation using the setpoint signal S. This corresponds to the usual understanding of modulation in the physical sense.

[0067] The generated modulated setpoint signal mS i is used in a step b. instead of the setpoint signal S i for operational control BmS i used. When operating using the modulated setpoint signal mS iAccording to step c, vibration signals V are continuously recorded. Several vibration sensors VS can be provided to record the vibration signals V, and the signals from the vibration sensors VS can be read in by the control unit executing the method. The vibration signals V are each representative of local vibrations of the platform and / or components of the propulsion system. The vibration sensors VS can, for example, be mounted in pairs and, for example, opposite one another on the platform of the propulsion system. In the case of a ship as a system driven by the propulsion system, vibration sensors can be mounted in pairs at the stern, bow and on the sides, for example, inside or outside in the ship's hull.

[0068] In step d., vibration spectra are then determined from the vibration signals V or vibration data, for example, by a sliding Fourier transform (FT) or a sliding fast Fourier transform (FFT). For the sliding determination in step d., a sliding window is used that has a predetermined window length that is smaller than a period of the modulation signal M. i .

[0069] In step d., remote frequency components from the respective vibration spectra are determined for each window of the sliding determination.

[0070] After determining the vibration spectra and frequency components, in step e., a cost function K is calculated for each of the windows for the respective frequency components, and the window for which the cost function has a minimum is determined.

[0071] The cost function for the frequency components is, in the sense of the underlying invention and in the usual technical sense, an objective function within the framework of an optimization task or optimization problem in which one or more technical parameters are optimized. In this respect, and merely for the sake of clarity, it should be mentioned that the component "costs" in the term "cost function" is not to be understood in the sense of monetary costs in business terms. The underlying invention relates in particular to the optimization or reduction or minimization of the noise signature, so that from a technical point of view the component "costs" refers to the noise signature to be optimized, whereby the "costs" are, for example, lower the smaller the noise signature.

[0072] In the sense of the invention, the optimization task is performed by minimizing the cost function, which means that the noise signature is optimized, reduced, or minimized. In other words, from a technical perspective, minimization should generally refer to the noise signature toward a reduction, in particular minimization, of the noise signature, even if, from a strictly mathematical perspective, a minimization task can also be realized by a maximization task using a different or opposite objective function, because the technical effect of minimization is identical in both cases. What this means is that minimization concerns the goal of reducing the noise signature and not necessarily individual mathematical operations for finding a minimum of a function.

[0073] The window assigned to the minimum corresponds to the window with the smallest noise signature among those considered in the sliding calculation. This means, in turn, that the window with the minimum cost function has, at least on average, the smallest noise signature compared to the other windows. Since the setpoint signal S i in each window with the modulation signal M i is modulated, the signal generated by the modulation signal M i in the window with minimal noise signature caused modulated change of the setpoint signal S i at least on average, compared to the other windows of the sliding analysis, leads to a reduction of the noise signature. Therefore, it can be assumed that the time average of the modulation signal M ifor the window assigned to the minimum, considered across all windows of the sliding analysis, leads to a reduction or optimization of the noise signature.

[0074] Accordingly, in step f. the time average value ZM of the modulation signal M i for the window assigned to the minimum, and in the following step g. the setpoint signal S is corrected i based on the time average ZM i , ie a corrected setpoint signal S' i The corrected setpoint signal S' i can be calculated, for example, by adding the time average ZM i to the setpoint signal S i The determined corrected setpoint signal S' i is then used instead of the setpoint signal S i used to control the operation of the drive system.

[0075] Steps a. to g. form an iteration stage of a continuous or iterative process for optimizing the noise signature, in which steps a. to g. are executed continuously and iteratively. If S i the setpoint signal in iteration step i (i = 1, 2, 3, ...), the corrected setpoint signal S' obtained in iteration step i i the setpoint signal S i+1 for the subsequent iteration step i+1, ie S i+1 = S' i , where in iteration step i+1 the modulated setpoint signal mS i+1 is determined (mSi +1 = mS' i ) etc. This iteration can be continued in particular as long as a setpoint signal SV is specified at the beginning of an iteration step, for example. i a setpoint signal specified by a higher-level control unit, e.g. concerning the relative rotor position, continues to be relevant or unchanged.

[0076] The iterative execution of steps a. to g. is described in Fig. 1 is represented by a solid arrow connecting step g. with step a.

[0077] The way in which the noise signature is optimized forms the structure of a control loop, as can be seen from the above explanations.

[0078] In real-life operation of the propulsion system, for example, in an eVSP or other marine propulsion system, it is common for the setpoint signal S to be specified by a control unit in the form of a setpoint signal specification SV. The setpoint signal specification SV can change depending on the desired control of the propulsion system. For example, in real-life operation of the eVSP, the setpoint signal specification SV can change regarding the relative rotor position.

[0079] For such cases, the method can provide that, before performing step a, a check is made to determine whether there has been a change in the setpoint signal specification SV, or whether there is a new setpoint signal specification SV. For example, if at the beginning of an iteration i there is a setpoint signal specification SV i which differs from the setpoint signal specification S i-1 of iteration step i-1, in step a. of the relevant iteration i the new setpoint signal SV i associated setpoint signal S(SV i ) instead of the corrected setpoint signal S' i-1 used. In Fig. 1, corresponding process steps when changing the setpoint signal specification are marked by dashed lines or borders.

[0080] Specifically, for example, after iteration i-1 and before iteration i is executed, it is checked whether the setpoint signal specification SV has changed, ie it is checked whether the condition SVi-1 = SV i is fulfilled. If this condition is fulfilled (“Yes”), ie there is no change in the setpoint signal specification, then in iteration i the setpoint signal S i the corrected setpoint signal S' determined in iteration step i-1 i-1 used (S i = S' i-1 ) and forms, in a sense, the new setpoint signal specification. If this condition is not met, ie there is a change in the setpoint signal specification SV, then in iteration step i the new or changed setpoint signal specification SV i associated setpoint signal S(SV i )used (S i = S(SV i )), which in Fig. 1 is indicated by step a*.

[0081] In steps a. and a* of the Fig. 1 is by naming the modulation function M i further indicated that the modulation of the setpoint signal S i Modulation function M to be used ican be determined anew for each iteration step. For example, as explained above, for the modulation function M i Functions are taken into account that depend on the respective valid setpoint signal S i or other variable operating or control parameters of the drive system, such as the speed of the rotors.

[0082] The addition of step a* to the procedure and the dashed-line procedure in Fig. 1 represents merely one variant of the noise signature optimization according to the invention. From the above, it becomes particularly clear that the proposed method is particularly applicable to drive systems in which the setpoint signal specification is constant or remains unchanged over a comparatively long period of time. With a constant or constant setpoint signal specification, the iterative optimization takes into account, for example, changing environmental influences and possibly changing local conditions for the drive system.

[0083] The iterative process makes it possible to achieve an optimization of the noise signature dynamically and during operation of the drive system, in particular on the basis of the actual operating and control data of the drive system (at least based on the respective valid setpoint signal and the vibration data as input data).

[0084] Fig. Figure 2 shows an example of a temporal sequence of a noise signature optimization method, where the horizontal axis indicates the time course or time t, and the index i (i=1, 2, 3...) represents the iteration. For better understanding, the process steps are divided into work blocks A, B, C, and D, also called blocks for short. Fig. 2 denotes t i (i = 1, 2, 3...) denotes the starting time of iteration i, and Δt1, Δt2 ... denote time intervals.

[0085] Block A concerns the execution of steps e. to g. in iteration step i-1, i.e., calculating the cost function, determining the minimum, calculating the time average, and correcting or adjusting the setpoint signal. The time interval for executing steps e. to g. corresponds to [t i - (t i-1 + Δt2)].

[0086] Block B concerns the determination of the modulation function M iin iteration step i, generating the modulated setpoint signal mS i and applying the modulated setpoint signal mS i at time t i + Δt2.

[0087] Block C concerns the acquisition of the acceleration values, corresponding to step c., and extends over the time interval [t i +Δt2; t i +Δt4].

[0088] Block D comprises the sliding determination of the vibration spectra and the determination of the frequency components for respective windows, corresponding to step d., in a time interval [t i +Δt3; t i +Δt4].

[0089] The following is a concrete example of a procedure according to Fig. 1 and Fig. 2. The example is based on the following prerequisites and boundary conditions: - the drive system comprises two eVSPs, each driven by an electric motor, which have the same speed on average or over the period under consideration; - the setpoint signal specification SV concerns the relative rotor position with setpoint signal S(SV); - the modulation function has the general form a*sin(b*t); - the modulation function has the period T Mi = N rev * 60 / n k , where N rev (>2) the number of revolutions of the eVSP in the time interval [t i-1 ; t i ], and n k the average speed of the eVSP; - a sliding Fourier transformation is performed with a sliding window length of FL = T Mi / N rev ; - each rotor of the eVSP has five blades (note: the number of blades is, as will be shown below, not an influencing factor in the calculation of the modulation signal or its parameterization); - the amplitude a of the modulation function depends on or is a function of the average speed n k of the two eVSP: a = a(n k ); - in the cost function K, the integer multiples of the blade frequencies or rotor blade frequencies are taken into account in a weighted sum of squares.

[0090] This means in particular that the period of the modulation function is proportional to the number of revolutions of the eVSP in the previous iteration interval, the period is inversely proportional to the average speed of the two eVSPs, the window length is proportional to the period of the modulation signal and the window length is inversely proportional to the average speed.

[0091] Under the above assumptions, the modulation function M i the form: Mi=a(ni)∗sin(2πni60Nrev(t−ti))

[0092] The period of the modulation signal is: TMi=Nrev∗60ni

[0093] The window length is: FL=TMiNrev=60ni

[0094] The cost function is: Kj=γ0a0,j2+γ1a1,j2+…

[0095] Where j = 1, 2, 3..., γ j are weighting factors and a0, a1 ... are vibration values ​​corresponding to a respective fundamental vibration a0 and multiples of the rotor blade frequencies a1, a2, .... The rotor blade frequency can be calculated as follows: n k / 60*(number of blades per rotor).

[0096] According to Fig. 2 begins evaluation of the vibration signals at t i + Δt3 using FFT with sliding window after the vibration values ​​starting from time t i + Δt2 for a window length FL. The sliding windows with window length FL are in Fig. 2 indicated by consecutively offset double arrows.

[0097] In the Fig. In the example shown in Figure 2, the time span between t i and t i + Δt4 of the period T Mi of the modulation signal, ie the recording of the vibration values ​​and the associated sliding evaluation extends over exactly one period T Mi , where Δt4 = T Mi is.

[0098] Fig. 3 shows schematically the process of the sliding evaluation of the vibration signals, where the Fig. Curve V shown in Figure 3 represents the continuously recorded vibration signals, which are evaluated based on a sliding fast Fourier transform (FFT). FFT0, FFT1, etc., denote the FFT evaluation in the window or window interval 0, 1, ....

[0099] Fig. 4 shows schematically an evaluation result of the vibration signals based on FFT, e.g. in window interval 1. The vibration spectrum shows exemplary signals at the fundamental frequency f R and at rotor blade frequencies 5f R , 10f R , 15f R and 20f R . The corresponding amplitudes a0, ... , a5 of the spectrum signals can be taken into account in the cost function as described above.

[0100] For the example of Fig. 2 to 4 results for a modulated setpoint signal mS to be used in iteration i i : mSi=SSVi(t)+Mi(t)+ΔSi−1corr+Sint

[0101] Here S svi the setpoint signal valid at the start time of iteration i according to setpoint signal specification SV i , M i (t) the modulation function determined for iteration step i, ΔSi−1corr the correction value for the setpoint signal from the previous iteration i-1 and S int further components of the setpoint signal, for example based on corrections from previous iterations.

[0102] Fig. Figure 5 schematically shows the structure of a ship platform 1 with two eVSPs 2, which are driven by electric motors (not shown in detail), e.g., PMSMs. Several vibration sensors 3, a total of six, are mounted on the ship platform 1. Each eVSP has five rotor blades 4. The rotor positions of the eVSPs are indicated by φ1 and φ2. The vibration sensors 3 are arranged symmetrically to each other and in pairs opposite each other. The rotor positions are specified or determined accordingly by a setpoint signal specification for the relative rotor position.

[0103] As can be clearly seen from the above explanations, for the Fig. 5, the system shown optimizes the noise signature according to a method proposed herein, in particular based on the results obtained in connection with Fig. 1 to 4 described embodiments.

[0104] Fig. Figure 6 schematically shows a block structure of a propulsion system for a ship with two eVSP 2 and associated control and monitoring components. The eVSP 2 are each driven by a permanent magnet synchronous motor PMSM, with corresponding speed controls 5 being provided to control the speed of the PMSM. The respective speeds can be implemented based on a thrust command, coming, for example, from the ship's bridge, for example, via one or more respective controllers 7. The thrust command is ultimately converted into a thrust 8.1 and 8.2 of the two eVSPs. The rotor positions of the eVSPs are shown in Fig. 6 denoted by φ1 and φ2.

[0105] In the control structure described above, a method designed according to an embodiment of the underlying invention can be implemented to optimize or reduce the noise signature.

[0106] Components for implementing such a procedure are in Fig.6 are highlighted by dashed-line rectangles. In particular, a control unit 8 can be provided, which optimizes the relative rotor position according to an embodiment of the method proposed herein. Input data 9.1 to 9.3 for the control unit are, in particular, acceleration values ​​(9.1), such as a = a1 ... a6 from a total of six vibration sensors 3, the rotor positions (9.2) φ1 and φ2, and the relative rotor position (9.3) Δφ according to a setpoint signal specification or according to a setpoint signal for the relative rotor position. Output data 10 of the control unit 8 can be, for example, correction values ​​relating to the relative rotor position, which can be supplied to the speed controllers 5 as additional input data using weighting factors 11.1 and 11.2, for example in the form of correction values. Preferably, the weighting factor 11.1 = ½ and the weighting factor 11.2 = -½, or vice versa.

[0107] By continuously iterating the relative rotor position correction method, the proposed method makes it possible to reduce, in particular to minimize, the noise signature.

[0108] The advantages of the proposed method are in particular that vibrations and noise generation on the ship - or generally on the platform - can be minimized, especially during ongoing operation ("online", dynamically or in real time). Determining the optimal relative rotor position - or generally the optimal setpoint signal - during ongoing operation enables high robustness and, for example, continuous operation without impairing the normal operation of the propulsion system. Furthermore, the method can be implemented for controlling the relative rotor position during ship operation without additional ship parameters - only the rotor position and vibration values ​​are required. Permanent, continuous control of the relative rotor position to the optimal operating point can be implemented, in particular with a minimal noise signature.

[0109] Furthermore, unlike the state of the art, no test drives or test runs of the drive system are required to determine the optimal relative rotor position, so that the optimization of the noise signature is not limited to specific operating modes. It is also possible to change the evaluation criteria used to optimize the noise signature depending on the operating point of the drive system or platform, and to selectively select and weight certain force components, vibration components, or force harmonics, which provides a comparably high level of flexibility. A further advantage is that the modulation signal can be fully adapted via software depending on the operating point, which also leads to a high level of flexibility and enables adaptability to numerous different systems and operating conditions and environments.

[0110] The proposed method has a wide range of applications and is suitable, for example, for marine propulsion systems, particularly eVSPs, thrusters (e.g., Voith Inline Thrusters or Voith Inline Thrusters), wind turbines, turbocompressors, etc. In the field of marine propulsion, the method is particularly suitable, but without limitation, for wind turbine maintenance vessels, research vessels, and harbor tugs. By optimizing the noise signature, it is possible to increase crew comfort, protect underwater wildlife, and counteract the interference with onboard active sonar systems. Reference symbol 1 ship platform 2 eVSP 3 Vibration sensor 4 rotor blades 5 Speed ​​control 6 Thrust specification 7 Control 8 Control unit 9 Input data 10 Initial data 11 Weighting factor ag procedural steps A ... D work blocks BmS operating control with modulated setpoint signal FT Fourier transform FFT fast Fourier transform FL window length K cost function M Modulation signal / modulation function S Setpoint signal S' corrected setpoint signal SV setpoint signal specification mS modulated setpoint signal T period (modulation function) t time V Vibration signal VS vibration sensor ZM time average φ rotor position / phase position QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited non-patent literature

[0000] Tom A. Smith et al, “Underwater radiated noise from marine vessels: A review of noise reduction methods and technology,” Ocean Engineering, Vol. 266, Part 1, Dec. 15, 2022

[0009] Synchrophasing - Enhanced vibration performance for twin propulsion vessels“, MAN Energy Solutions, 5510-0247-00 Mar 2020, available at: https: / / www.man-es.com / docs / defaultsource / marine / tools / synchrophasing.pdf?sfvrsn=3c464a88_14

[0015]

Claims

[1] Computer-implemented method for reducing the noise signature in a drive system mounted on a platform (1), which comprises at least one driving and one driven (2) component, wherein a control signal with at least one setpoint signal (S) is provided for controlling the operation of the driving or the driven (2) component according to an operating requirement, wherein the method is carried out iteratively, and a respective iteration step (i) comprises the following steps: a. generating a modulated setpoint signal (mS) by modulating the setpoint signal (S) with a modulation signal that can be described by a periodic function (M), such that a time average of the modulated setpoint signal (mS) corresponds to a time average of the setpoint signal (S), b. Applying the modulated setpoint signal (mS) instead of the setpoint signal (S) to control the operation of the driving or driven (2) component, c. during the application of the modulated setpoint signal (mS), continuous detection of vibration signals (V) representative of local vibrations of the platform (1) and / or the component by at least one vibration sensor (VS, 3), d. sliding determination of vibration spectra with a sliding window from the vibration signals (V), wherein the sliding window has a predetermined window length (FL) which is smaller than a period (T) of the modulation signal (M), and determining frequency components (f R , 5f R , ...) for each window from respective vibration spectra, e. for each of the windows, calculating a cost function (K) for the respective frequency components (f R , 5f R, ...), and determining the window for which the cost function (K) has a minimum, f. Calculating a time average (ZM) of the modulation signal (M) for the window associated with the minimum, and g. Correcting the setpoint signal (S) based on the time average (ZM) and applying the corrected setpoint signal (S') instead of the setpoint signal (S). [2] Method according to claim 1, wherein the period (T) of the modulation signal (M) and / or an amplitude (a) of the modulation signal (M) varies for temporally successive iteration steps, and is calculated, preferably, as a function of the setpoint signal and / or a rotational speed applied, determined, measured or estimated in a preceding or respective iteration step. [3] Method according to one of claims 1 or 2, wherein the window length (FL) is proportional to the period (T) of the modulation signal (M) and / or wherein the window length (FL) is inversely proportional to a number of revolutions of the driving or driven (2) component per period (T) and / or wherein the window length (FL) is inversely proportional to a rotational speed of the driving or driven (2) component in the respective iteration step. [4] Method according to one of the preceding claims, wherein the modulation signal (M) can be described by a sine or cosine function or is given by a sine or cosine function. [5] Method according to one of the preceding claims, wherein the frequency components (f R , 5f R , ...) in step d. a fundamental frequency (f R ) and harmonic frequencies (5f R , 10f R ...) the fundamental frequency (f R), and wherein in step e. the cost function (K) is determined as a weighted sum, in particular a weighted quadratic sum, of the fundamental frequency (f R ) and the harmonics (5f R , 10f R ...) is calculated. [6] Method according to one of the preceding claims, wherein the at least one driving or driven (2) component comprises a rotor (2), and the setpoint signal preferably specifies a rotational speed, a torque, a power and / or a phase position (φ) of the rotor (2). [7] Method according to claim 6, wherein h. the period (T) of the modulation signal (M) is inversely proportional to the speed of the rotor (2), i. the period (T) of the modulation signal (M) is proportional to the number of revolutions of the rotor (2) during a period (T), and / or j. the window length (FL) in an iteration step is inversely proportional to the number of revolutions of the rotor (2) during a period (T) of a respective / previous iteration step. [8] Method according to one of claims 1 to 7, wherein the drive system comprises at least two driving or driven (2) components, each having a rotor (2), wherein each rotor (2) preferably comprises at least two rotor blades (4), and wherein the setpoint signal specifies a relative phase position for at least two of the rotors (2). [9] Method according to one of claims 1 to 8, wherein the drive system comprises at least one, preferably two, driving components which are designed as a motor, in particular an electric motor, two, in particular exactly two, rotors (2) driven by the driving component(s) as driven components, wherein each rotor (2) comprises a plurality of rotor blades (4) connected thereto, wherein the drive system is preferably designed as a ship drive, in particular comprising two or more electric motor-driven Voith-Schneider propellers (2) as driven components, wherein the setpoint signal (S) specifies a relative rotor position of at least two, preferably exactly two, of the rotors (2), wherein for each time window an amplitude (a) and the period (T) of the modulation signal (M) are determined, wherein a. the amplitude (a) of the modulation signal (M) is a function of the mean value of the rotational speeds of the rotors (2) in the respective / previous iteration step, b. the period (T) is inversely proportional to the average speed of the respective rotor (2) in the respective / previous iteration step and proportional to the number of revolutions of the rotor blades (4) per period (T) of the respective / previous iteration step and / or is inversely proportional to the average speed of the respective rotor (2) in the respective / previous iteration step, c. the window length (FL) is proportional to the period (T) and inversely proportional to the number of revolutions of the rotor blades (4) per period (T) of the respective / previous iteration step, and d. as frequency components (f R , 5f R , ...) a fundamental frequency (f R) and integer multiples of the fundamental frequency (f R ) should be taken into account. [10] A data processing system (8) comprising means for carrying out the method steps according to one of claims 1 to 9, wherein the means are configured to receive or determine the setpoint signal (S) and the vibration signals (V) of the at least one vibration sensor (VS, 3) and to provide the corrected setpoint signal (S') for controlling the operation of the drive system or to control the drive system on the basis of the corrected setpoint signal (S'). [11] System according to claim 10, further comprising the at least one vibration sensor (VS, 3), preferably a plurality of vibration sensors (VS, 3). [12] A computer program adapted to carry out a method according to any one of claims 1 to 9 when executed on a data processing system (8) according to claim 10. [13] A computer-readable medium comprising instructions which, when executed by a data processing system (8) according to claim 10, cause the system (8) to carry out the method steps of a method according to any one of claims 1 to 9. [14] Drive system, in particular for watercraft, comprising at least one, preferably two, motors, preferably designed as an electric motor, as driving components, at least one, preferably two, rotors (2) as driven components and a plurality of rotor blades (4) coupled to a respective rotor (2), one or more vibration sensors (VS, 3) and a data processing system (8) according to claim 10, wherein the driven component as a rotor (2) with the rotor blades (4) coupled thereto each preferably comprises an electrically driven Voith-Schneider rotor. [15] Watercraft comprising a platform (1) with at least two rotors (2) attached thereto, each having a plurality of rotor blades (4), in particular according to a Voith-Schneider rotor, one or more vibration sensors (VS, 3) coupled to the platform (1) and a control system for controlling the operation of the rotors (2), wherein the control system comprises a data processing system (8) according to claim 10.

Images